You are here

  • Home
  • Archive
  • Volume 101, Issue 6
  • Neurodevelopmental outcome at 2 years in offspring of women randomised to metformin or insulin treatment for gestational diabetes

Article Text

Article menu

Download PDFPDF

Original article
Neurodevelopmental outcome at 2 years in offspring of women randomised to metformin or insulin treatment for gestational diabetes
  1. Trecia A Wouldes1,
  2. Malcolm Battin2,
  3. Suzette Coat3,
  4. Elaine C Rush4,
  5. William M Hague5,
  6. Janet A Rowan6
  1. 1Department of Psychological Medicine, Faculty of Medical and Health Sciences, Auckland, New Zealand
  2. 2New Born Services, Auckland City Hospital, Auckland, New Zealand
  3. 3Robinson Research Institute, The University of Adelaide, North Adelaide, South Australia
  4. 4Centre for Child Health Research, Auckland University of Technology, Auckland, New Zealand
  5. 5Department of Obstetrics, Women and Children's Hospital, University of Adelaide, North Adelaide, South Australia
  6. 6Department of Obstetrics, National Women's Health, Auckland City Hospital, Auckland, New Zealand
  1. Correspondence to Professor Trecia A Wouldes, Department of Psychological Medicine, Faculty of Medical and Health Sciences, Private Bag 92019, Auckland 1142, New Zealand; t.wouldes{at}auckland.ac.nz

Abstract

Objective Gestational diabetes mellitus (GDM) is a common complication of pregnancy and is increasingly being treated with metformin that crosses the placenta rather than insulin, which does not. This study seeks to examine the neurodevelopment of offspring of women treated with metformin or insulin for GDM.

Design We performed a prospective follow-up study of children whose mothers had been randomly assigned at 20–33 weeks gestation to treatment with metformin or insulin for GDM. Of the 211 children followed up at 2 years, 128 were from Auckland, New Zealand (64 metformin vs 64 insulin), and 83 from Adelaide, Australia (39 metformin vs 49 insulin). Neurodevelopment was examined with the Bayley Scales of Infant Development V.2 mental development index (MDI) and psychomotor development index (PDI). Clinical and demographic background characteristics were obtained at enrolment, birth and follow-up.

Results No significant differences were found between treatment groups in clinical or demographic characteristics. The MDI and PDI composite scores were tested with general linear models. No significant differences were found between metformin and insulin, respectively, in New Zealand (MDI, M=83.6 vs 86.9 and PDI, M=83.4 vs M=85.2) or Australia (MDI, M=102.5 vs M=98.4 and PDI, M=105.6 vs M=99.9) and no interactions observed. The differences in neurodevelopmental outcomes between the Auckland and Adelaide cohorts were explained by parental ethnicity, infant birth weight >4000 g, neonatal hypoglycaemia, maternal glycaemia and smoking in the household.

Conclusions This study provides additional data supporting the safety of metformin in the treatment of GDM.

Trial registration number ACTRN 12605000311651.

  • Neurodevelopment
  • Outcomes research
  • Gestational Diabetes Mellitus
  • Metformin

https://doi.org/10.1136/archdischild-2015-309602

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    What is already known on this topic?

    • Gestational diabetes mellitus (GDM) that cannot be managed with lifestyle changes has typically been treated with insulin, but increasingly metformin, which crosses the placenta, is being used.

    • Treatment of GDM improves pregnancy and birth outcomes; however, very little evidence exists regarding the effects of treatment on later neurodevelopment of the offspring.

    What this study adds?

    • Neurodevelopmental outcomes at 2 years of age are similar between children born to mothers who were treated with metformin compared to those treated with insulin for GDM.

    • Lower scores on standardised measures of neurodevelopment were predominantly associated with parental self-identified ethnicity, smoking in the household and clinical outcomes at birth unrelated to treatment status.

    Gestational diabetes mellitus (GDM) is a common complication of pregnancy, with increasing rates worldwide.1–4 Offspring of mothers with GDM have been reported to have deficits in fine and gross motor function,5–9 expressive language development, verbal and general IQ,6 ,10–14 lower rates of educational attainment,7 ,12 more attention and behavioural problems,6 ,10 and, more recently, a risk of autism.15 Other studies report no deficits in cognitive function.16–18 However, studies often did not distinguish between pre-existing diabetes and GDM,8 ,12 ,14 report maternal metabolic control or whether GDM was treated and how.10–12 ,14

    Treatment of GDM improves pregnancy outcomes.19 ,20 Until recently, treatment has been with lifestyle adjustment, and where necessary, insulin, which does not cross the placenta. Metformin, which does cross the placenta, is increasingly being used as it has been shown to be a safe and effective alternative to insulin with respect to pregnancy outcomes.21–24 Yet, there are few data on the neurodevelopment of offspring born to mothers with GDM treated with insulin compared with those treated with metformin. One randomised study reported no difference in motor, social or linguistic outcomes in 97 children (47 metformin, 50 insulin) assessed by questionnaire at 18 months. Another observational study assessed serial neurodevelopment over 18 months in infants where the mothers had continued metformin prescribed for the treatment of polycystic ovary syndrome (PCOS) and reported no evidence of delay.25

    The largest study, to date, demonstrating the effectiveness of metformin compared with insulin to treat GDM has been the Metformin in Gestational diabetes (MiG) trial,24 in which 733 women with GDM requiring pharmacotherapy were randomised to insulin or metformin (with additional insulin as necessary) to achieve glucose targets. The offspring of these women are being followed, and studies of their body composition at 2 years of age have already been reported.26 This study compares their neurodevelopmental outcomes at the same time point.

    Methods

    Participants were offspring of 211 women who had been enrolled in the MiG study from two sites in Auckland, New Zealand (128 children in total), and one site in Adelaide, Australia (83 children) (figure 1). Women, who had earlier consented to further follow-up, were contacted by telephone after their child's second birthday and the protocol for the study was explained. In New Zealand, written consent, a maternal interview and anthropometry measures of the mother and child were collected at a home visit and a further follow-up visit was arranged within 2 weeks to attend the Liggins Institute Clinical Research Unit for neurodevelopmental assessment, physical examination and dual-energy X-ray absorptiometry measurement as previously described.27 In Adelaide, written consent and all interviews and assessments were completed at the Women's and Children's Hospital. Registration of the MiG randomised controlled trial had been obtained prior to its initiation through the Australian New Zealand Clinical Trials Registry (ACTRN 12605000311651).

    Figure 1
    Figure 1

    Offspring followed up from the Metformin in Gestational diabetes (MiG) trial. DEXA, dual-energy X-ray absorptiometry; GDM, gestational diabetes mellitus.

    Demographics and clinical characteristics of the mother and child

    Self-identified ethnicity and level of education were obtained from parents at enrolment in the MiG study. Perinatal data for the mother and child were obtained from the MiG data set including infant birth weight, length, head circumference, gestation and incidence of neonatal blood glucose <2.6 or 1.6 mmol/L and maternal HbA1c during pregnancy.24 ,27 Follow-up questionnaires, administered by trained researchers, provided the following: (1) socioeconomic status (SES) of the family (any government benefit, home ownership and level of parents’ education), ethnicity and whether English was the first language; (2) number of children in the family; (3) household alcohol and/or tobacco use; (4) maternal depression (yes/no) or on medication for depression (yes/no); and (5) history of serious child health problems since birth (yes/no)—determined by asking parents about any serious illnesses at the follow-up appointment.

    Measure of neurodevelopment

    Examiners blinded to treatment status administered the Bayley Scales of Infant Development V.2 (BSID-II)28 at the 2-year follow-up. Age of administration was corrected for prematurity for those children born at <37 weeks. The BSID-II is a standardised measure of neurodevelopment comprised of three components: the Mental Developmental Index (MDI), the Psychomotor Developmental Index and the Behaviour Rating Scale (BRS). The MDI and psychomotor development index (PDI) provide developmental quotients with a mean (M) of 100 and an SD of ±15. The BRS is completed by the examiner at the end of the administration of the MDI and PDI and includes subscales covering orientation, emotional regulation, motor quality plus a total score.

    In the New Zealand sample (N=95), a structured physical examination was performed by a paediatrician and the findings recorded as a separate item in the database. Only 10 of the Australian children underwent a formal physical examination.

    Statistical analysis

    Analyses were performed with the IBM Statistical Package for Social Sciences V.21. Significance was set at p<0.05.

    Analyses of variance and χ2 statistics were used to compare the demographic and clinical characteristics of the metformin and insulin groups in each country at birth and at follow-up.

    BSID-II scores on the MDI and PDI were tested with general linear models, for treatment group (metformin vs insulin), country (New Zealand vs Australia) and the interaction between treatment and country. χ2 statistics were used to test the association by treatment and country of the Bayley BRS and the subscales: orientation, emotion regulation and motor quality. These were categorised by the clinical cut-offs for non-optimal, questionable and within normal limits.28 Bonferroni correction for multiple comparisons was used where appropriate.

    After observing differences in neurodevelopmental outcomes between the two centres, we conducted further analyses to explore important associated factors. Multivariate regression models were fitted in which infant outcomes on the Bayley MDI and PDI were modelled as a linear function of treatment in relation to the following: social characteristics at birth and at follow-up, maternal glycaemic control, neonatal characteristics, age at follow-up and incidence of any major childhood illnesses. The choice of social and clinical covariates was based on both theoretical and analytical considerations and as not being highly correlated. Dichotomous variables were coded 0 (yes) or 1 (no). Models were fitted using forward and backward elimination (<0.05) of the above variables to obtain the best explanation of variation in scores.

    Results

    Of the 128 New Zealand participants studied, the mothers of 64 were treated with metformin and the mothers of 64 were treated with insulin. Of the 83 Australian participants, the mothers of 39 were treated with metformin and the mothers of 44 were treated with insulin. No differences were found between treatment groups within the New Zealand or Australian cohorts with regard to self-identified ethnicity or educational attainment (table 1). Overall, the Australian cohort had predominantly mothers and fathers of European descent (92% and 94%, respectively). The New Zealand cohort was more diverse: fewer than half of the mothers and fathers were European (45% and 46%), Indian mothers and fathers made up the next largest group (16% and 18%), followed by Pacific Islands mothers and fathers (13% and 18%). The remaining cohort included a small proportion of Chinese, other Asian, Maori and other ethnicities. Due to this ethnic diversity, English was the second language for 20% of the families in the New Zealand cohort compared with 6% in the Australian. Over 80% of mothers and fathers in both cohorts had either a secondary or tertiary education.

    View this table:
    Table 1

    Comparison of maternal and paternal characteristics at randomisation and at follow-up by treatment and by country

    A larger proportion of children of European or white ethnicity was available for the neurodevelopmental follow-up at 2 years compared with the original MiG population (metformin 68% vs 44%, p<0.001; insulin 58% vs 43%, p<0.005), whereas there was a smaller proportion of children of Pacific or Polynesian ethnicity similarly compared (metformin 7% vs 22%, p<0.001; insulin 13% vs 21%, p<0.060).

    At recruitment into the MiG trial, women in the New Zealand cohort had higher mean HbA1c levels than the Australian cohort, 5.8 (0.1)% vs 5.3 (0.1)%, p<001, reflecting higher glucose levels prior to intervention.

    At follow-up, the characteristics of the family environment and SES were similar among treatment groups across most measures in both New Zealand and Australian cohorts (table 1). However, more New Zealand fathers in the insulin group were living with the mother and child at follow-up than in the metformin group (91% vs 78%, p=0.05), and more households in the Australian cohort consumed alcohol than in the New Zealand cohort (58% vs 35%, p=0.004).

    No differences in perinatal outcomes were found between treatment groups and between the New Zealand and Australian cohorts, except that in the New Zealand cohort birth length in the insulin group was longer than that in the metformin group (table 2).

    View this table:
    Table 2

    Comparison of offspring characteristics at birth and at follow-up by treatment and country

    At follow-up, no differences were detected in the proportions of children who had serious health outcomes both between treatment groups and between cohorts.

    There was a difference between the New Zealand and Australian cohorts in respect of the age at neurodevelopmental assessment (mean age in months (SD) 27.40 (2.41) vs 33.36 (1.68), p<0.001, respectively).

    In New Zealand, physical examination found the majority of children to be normal. One child required a hearing aid and in one child it was difficult to elicit tendon reflexes.

    Four children were identified as having speech delay (two metformin and two insulin) and two were referred for follow-up, one diagnosed with autism and one with a seizure disorder (A3243 mitochondrial mutation).

    In Australia, where childhood conditions were collected from parental interview, 10 children (4 metformin, 6 insulin) were reported to have speech delay. In the insulin arm, one child required hearing aids, one child had been born with tracheo-oesophageal fistula and cleft palate, one had been diagnosed with Angelman syndrome, one with neurofibromatosis (type 1) and two with Asperger's syndrome.

    Neurodevelopmental outcomes by treatment and country are presented in table 3. In both cohorts, cognitive (MDI) and motor (PDI) quotients in the metformin group were similar to those in the insulin group. There was no interaction between sites, but there were significant differences in MDI (M=85.2, SD=15.6 vs M=100.3, SD=16.5) and PDI (M=84.3, SD=14.3 vs M=102.4, SD=15.2) scores between the New Zealand and Australian cohorts, respectively. Mean scores for the New Zealand cohort in both groups were approximately 1.0 SD below the BSID-II standardised mean (100±15) on both the MDI and PDI.

    View this table:
    Table 3

    Comparison of neurodevelopmental outcomes by treatment and by country

    Results in tables 1 and 2 raise the possibility that the observed differences between New Zealand and Australian cohorts in the MDI and PDI (table 3) could reflect the effect of social and/or clinical factors at birth and at follow-up of the individual cohorts. To explore this question, we fitted a series of standard regression models to the combined New Zealand and Australian cohorts (table 4). These analyses showed that being born to mothers of Pacific or Indian descent, having English as a second language, living in a household where adults smoked and having a birth weight >4000 g were all independently associated with lower scores on the MDI whereas being of white or Caucasian descent was associated with higher scores explaining 21% of the variability (F=11.13, p<0.001, R2=0.23, adjusted R2=0.21). Lower scores on the PDI were also associated with being born to mothers of Pacific or Indian descent, higher maternal HbA1c during pregnancy and having two or more recorded glucose values <2.6 mmol/L at birth. These factors explained 15% of the variability in the composite PDI scores (F=7.49, p<0.001, R2=0.18, adjusted R2=0.15).

    View this table:
    Table 4

    Regression analyses exploring associations between maternal and infant characteristics and neurodevelopmental outcomes

    Discussion

    This study provides strong evidence that there is no difference between the neurodevelopment of children at age 2 years whose mothers had received metformin during pregnancy compared with those whose mothers received insulin to treat their GDM. Global cognitive and motor development at the 2-year follow-up of the MiG trial offspring did not differ between treatment groups in either New Zealand or Australia. These results have important clinical implications given the rise in the incidence of GDM,2 and the emerging evidence that GDM is a potentially significant risk factor for developmental delay and disorders.5–15

    These findings are consistent with two earlier studies that used questionnaires to assess social, motor25 ,29 and linguistic29 outcomes over 18 months born to women who received metformin treatment during pregnancy to treat PCOS25 and women who were randomly assigned to receive metformin or insulin to treat GDM.29

    Two earlier studies have shown outcomes on the BSID-II and other measures of neurodevelopment in early childhood among offspring born to mothers with GDM that are also consistent with our results. The first found offspring of diabetic women had mean MDI scores corrected for SES and racial or ethnic origin at 2 years that were not different from those of the offspring of non-diabetic women (M=90 vs M=89). Children between 3 and 5 years of age had similar outcomes on the Stanford-Binet Intelligence Scale (M=91 vs M=92).30 The second compared offspring of diabetic women with offspring of women with uncomplicated pregnancies at 12 months of age.14 There was a difference in MDI (M=103 and M=95) scores between diabetic and non-diabetic groups, but not PDI scores (M=101 vs M=102). However, the sample was predominantly white (91%) born to families of middle to high SES and it is not clear whether the women had pre-gestational or gestational diabetes, or how it was managed.

    While we found no differences between treatment groups in neurodevelopmental outcomes, we did find significantly lower scores on both cognitive and motor development in New Zealand offspring compared with Australian offspring. Subsequent analyses found lower scores were predominantly associated with maternal ethnicity (Pacific or Indian) or lifestyle factors, and higher scores with being Caucasian. In addition, children with a birth weight >4000 g were more likely to have poorer cognitive development and children who had two or more occurrences of neonatal glucose values <2.6 mmol/L and offspring of mothers with higher glucose levels were more likely to have poorer motor development regardless of treatment.

    The major strengths of our study are the inclusion of an ethnically diverse, yet well-matched sample of women from two different countries, and the collection of environmental data at birth and at follow-up that explains the differences found between countries and within children. The low follow-up rate from the original trial, and the lower proportion of Polynesian children available at follow-up is a potential limitation. However, a similar proportion of Pacific Island families were available for follow-up in each treatment group; therefore, this is not likely to affect our results.

    Conclusion

    This is strong evidence that the neurodevelopment of offspring of women with GDM at 2 years is not affected by treatment with metformin, which can cross the placenta, compared with treatment with insulin. Our results are largely in line with other studies of neurodevelopment in children born to women with diabetes. However, the literature on the long-term neurodevelopment of such children is inconclusive. In early childhood, differences may be subtle, but with increasing age come the demands of higher order, more complex thought processes such as working memory, inhibition and, in adolescence, hypothesis testing; therefore, it remains to be seen whether these findings persist.

    Acknowledgments

    The authors would like to acknowledge the additional people who performed clinical and neurodevelopmental outcomes throughout the study. Natasha Johnson and Melanie Stephenson, Department of Psychological Medicine; Aida Siegers, Jenny Rafferty and Mariam Buksh from National Women's Health, Auckland, New Zealand; Jewel Wen, Neil Snowling, Jennifer Crowley and Sarah Bristow from the Auckland University of Technology, New Zealand.

    References

    1. ↵
      American Diabetes Association. Gestational Diabetes Mellitus. Diabetes Care 2003;26(Suppl 1):S103–5. doi:10.2337/diacare.26.2007.S103
      OpenUrlCrossRefPubMed
    2. ↵
      1. Reece EA,
      2. Leguizamón G,
      3. Wiznitzer A
      . Gestational diabetes: the need for a common ground. Lancet 2009;373:1789–97. doi:10.1016/S0140-6736(09)60515-8
      OpenUrlCrossRefPubMedWeb of Science
    3. ↵
      1. Metzger BE,
      2. Lowe LP,
      3. Dyer AR
      , et al. Hapo Study Cooperative Research Group. Hyperglycemia and adverse pregnancy outcomes. N Engl J Med 2008;358:1991–2002. doi:10.1056/NEJMoa0707943
      OpenUrlCrossRefPubMedWeb of Science
    4. ↵
      1. Vohr BR,
      2. Boney CM
      . Gestational diabetes: the forerunner for the development of maternal and childhood obesity and metabolic syndrome? J Matern Fetal Neonatal Med 2008;21:149–57. doi:10.1080/14767050801929430
      OpenUrlCrossRefPubMedWeb of Science
    5. ↵
      1. Silverman BL,
      2. Rizzo TA,
      3. Cho NH, et al
      . Long-term effects of the intrauterine environment. The Northwestern University Diabetes in Pregnancy Center. Diabetes Care 1998;21(Suppl 2):B142–9.
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Ornoy A,
      2. Ratzon N,
      3. Greenbaum C, et al
      . School-age children born to diabetic mothers and to mothers with gestational diabetes exhibit a high rate of inattention and fine and gross motor impairment. J Pediatr Endocrinol 2001;14(Suppl 1):681–9. doi:10.1515/JPEM.2001.14.S1.681
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Ornoy A,
      2. Wolf A,
      3. Ratzon N, et al
      . Neurodevelopmental outcome at early school age of children born to mothers with gestational diabetes. Arch Dis Child Fetal Neonatal Ed 1999;81:F10–14. doi:10.1136/fn.81.1.F10
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Ratzon N,
      2. Greenbaum C,
      3. Dulitzky M, et al
      . Comparison of the motor development of school-age children born to mothers with and without diabetes mellitus. Phys Occup Ther Pediatr 2000;20:43–57. doi:10.1080/J006v20n01_04
      OpenUrlCrossRefPubMed
    9. ↵
      1. Rizzo TA,
      2. Dooley SL,
      3. Metzger BE, et al
      . Prenatal and perinatal influences on long-term psychomotor development in offspring of diabetic mothers. Am J Obstet Gynecol 1995;173:1753–8. doi:10.1016/0002-9378(95)90422-0
      OpenUrlCrossRefPubMedWeb of Science
    10. ↵
      1. Nomura Y,
      2. Marks DJ,
      3. Grossman B, et al
      . Exposure to gestational diabetes mellitus and low socioeconomic status: effects on neurocognitive development and risk of attention-deficit/hyperactivity disorder in offspring. Arch Pediatr Adolesc Med 2012;166:337–43. doi:10.1001/archpediatrics.2011.784
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Dionne G,
      2. Boivin M,
      3. Séguin JR, et al
      . Gestational diabetes hinders language development in offspring. Pediatrics 2008;122:e1073–9. doi:10.1542/peds.2007-3028
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Fraser A,
      2. Almqvist C,
      3. Larsson H, et al
      . Maternal diabetes in pregnancy and offspring cognitive ability: sibling study with 723,775 men from 579,857 families. Diabetologia 2014;57:102–9. doi:10.1007/s00125-013-3065-z
      OpenUrlCrossRefPubMed
    13. ↵
      1. Fraser A,
      2. Nelson SM,
      3. Macdonald-Wallis C, et al
      . Associations of existing diabetes, gestational diabetes, and glycosuria with offspring IQ and educational attainment: the Avon Longitudinal Study of Parents and Children. Exp Diabetes Res 2012;2012:963735. doi:10.1155/2012/963735
      OpenUrlPubMed
    14. ↵
      1. DeBoer T,
      2. Wewerka S,
      3. Bauer PJ, et al
      . Explicit memory performance in infants of diabetic mothers at 1 year of age. Dev Med Child Neurol 2005;47:525–31. doi:10.1017/S0012162205001039
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      1. Xiang AH,
      2. Wang X,
      3. Martinez MP, et al
      . Association of maternal diabetes with autism in offspring. Jama 2015;313:1425–34. doi:10.1001/jama.2015.2707
      OpenUrlCrossRefPubMed
    16. ↵
      1. Ornoy A,
      2. Ratzon N,
      3. Greenbaum C, et al
      . Neurobehaviour of school age children born to diabetic mothers. Arch Dis Child Fetal Neonatal Ed 1998;79:F94–9. doi:10.1136/fn.79.2.F94
      OpenUrlAbstract/FREE Full Text
    17. ↵
      1. Ornoy A
      . Growth and neurodevelopmental outcome of children born to mothers with pregestational and gestational diabetes. Pediatr Endocrinol Rev 2005;3:104–13.
      OpenUrlPubMed
    18. ↵
      1. Veena SR,
      2. Krishnaveni GV,
      3. Srinivasan K, et al
      . Childhood cognitive ability: relationship to gestational diabetes mellitus in India. Diabetologia 2010; 53:2134–8. doi:10.1007/s00125-010-1847-0
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      1. Landon MB,
      2. Spong CY,
      3. Thom E, et al
      . A multicenter, randomized trial of treatment for mild gestational diabetes. N Engl J Med 2009;361:1339–48. doi:10.1056/NEJMoa0902430
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Crowther CA,
      2. Hiller JE,
      3. Moss JR, et al
      . Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med 2005;352:2477–86. doi:10.1056/NEJMoa042973
      OpenUrlCrossRefPubMedWeb of Science
    21. ↵
      1. Gui J,
      2. Liu Q,
      3. Feng L
      . Metformin vs insulin in the management of gestational diabetes: a meta-analysis. PLoS ONE 2013;8:e64585. doi:10.1371/journal.pone.0064585
      OpenUrlCrossRefPubMed
    22. ↵
      1. Su DF,
      2. Wang XY
      . Metformin vs insulin in the management of gestational diabetes: a systematic review and meta-analysis. Diabetes Res Clin Pract 2014;104:353–7. doi:10.1016/j.diabres.2013.12.056
      OpenUrlCrossRefPubMed
    23. ↵
      1. Dhulkotia JS,
      2. Ola B,
      3. Fraser R, et al
      . Oral hypoglycemic agents vs insulin in management of gestational diabetes: a systematic review and metaanalysis. Am J Obstet Gynecol 2010;203:457.e1–9. doi:10.1016/j.ajog.2010.06.044
      OpenUrlCrossRefPubMed
    24. ↵
      1. Rowan JA,
      2. Hague WM,
      3. Gao W, et al
      . Metformin versus insulin for the treatment of gestational diabetes. N Engl J Med 2008;358:2003–15. doi:10.1056/NEJMoa0707193
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Glueck CJ,
      2. Goldenberg N,
      3. Pranikoff J, et al
      . Height, weight, and motor-social development during the first 18 months of life in 126 infants born to 109 mothers with polycystic ovary syndrome who conceived on and continued metformin through pregnancy. Hum Reprod 2004;19:1323–30. doi:10.1093/humrep/deh263
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Rowan JA,
      2. Rush EC,
      3. Obolonkin V, et al
      . Metformin in gestational diabetes: the offspring follow-up (MiG TOFU): body composition at 2 years of age. Diabetes Care 2011;34:2279–84. doi:10.2337/dc11-0660
      OpenUrlAbstract/FREE Full Text
    27. ↵
      1. Rowan J,
      2. Rush EC,
      3. Obolonkin V, et al
      . Metformin in gestational diabetes: the Offspring Follow-up (MiG TOFU). Diabetes Care 2011;34:2279–84. doi:10.2337/dc11-0660
      OpenUrlAbstract/FREE Full Text
    28. ↵
      1. Bayley N
      . Bayley scales of infant development. 2nd edn. San Antonio, TX: The Psychologial Corporation, 1993.
    29. ↵
      1. Ijäs H,
      2. V ääräsmäki M,
      3. Saarela T, et al
      . A follow-up of a randomised study of metformin and insulin in gestational diabetes mellitus: growth and development of the children at the age of 18 months. Bjog 2015;122:994–1000. doi:10.1111/1471-0528.12964
      OpenUrlCrossRefPubMed
    30. ↵
      1. Rizzo TA,
      2. Ogata ES,
      3. Dooley SL, et al
      . Perinatal complications and cognitive development in 2- to 5-year-old children of diabetic mothers. Am J Obstet Gynecol 1994;171:706–13. doi:10.1016/0002-9378(94)90086-8
      OpenUrlCrossRefPubMedWeb of Science
    View Abstract

    Footnotes

    • Contributors All authors contributed to the design of the MiG TOFU study. TAW, MB and SC designed the neurodevelopmental follow-up. TAW and MB analysed and interpreted the data with JAR. TAW wrote a first draft of the manuscript, and all co-authors critically evaluated and suggested revisions to the manuscript and approved the final submission.

    • Funding This study was supported by funding from the Health Research Council, New Zealand; the Auckland Medical Research Foundation; the Evelyn Bond Trust, Auckland; and the National Health and Medical Research Council, Australia.

    • Competing interests None declared.

    • Patient consent Obtained.

    • Ethics approval Ethical approval for the study was obtained at each site.

    • Provenance and peer review Not commissioned; externally peer reviewed.